|Hsu, Wan-Ling - UNIV OF NEBRASKA-LINCOLN|
|Johnson, Rodger - UNIV OF NEBRASKA-LINCOLN|
|Van Vleck, Lloyd|
Submitted to: Journal of Animal Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: November 11, 2009
Publication Date: March 1, 2010
Citation: Hsu, W.L., Johnson, R.K., Van Vleck, L.D. 2010. Effect of Pen Mates on Growth, Backfat Depth, and Longissimus Muscle Area of Swine. Journal of Animal Science. 88(3):895-902. Interpretive Summary: Pigs are often tested for gain in feeding trials with relatively few pigs per pen (e.g., 8-15). A common perception is that the gain of a pig may be affected (negatively or positively) by the other pigs in the pen. Such effects have been variously named penmate, competition, associative, and social effects. Two important questions are 1) whether penmate genetic effects on gain should be included in a genetic selection index with direct genetic value for gain and 2) whether penmate effects interfere with prediction of direct genetic value that in selected animals would be expressed in commercial feedlots with many more pigs per pen whereas the penmate genetic effects would not be expected to be the same as in the feeding trials. The answers to both questions depend on how much variation in gain is accounted for by penmate effects. This variation is usually designated as the component of variance due to penmate effects. This component of variance can be contrasted to the component of variance due to direct genetic effects. With extremely small estimates of variance of penmate genetic effects, such effects could be ignored. Small estimates may be important because of the assumption that the penmate effect of one pig affects the gain of each of its penmates. Thus in a genetic selection index for gain, the estimate of the pig’s penmate genetic value would be weighted by the number of its penmates and the estimate of the pig’s direct genetic value would have a weight of one. Whether penmate effects in pens with few pigs would be the same as with many more pigs under commercial conditions has not been studied. This study of gains on test of 3,524 pigs with 9-11 per pen from experimental lines at the University of Nebraska did not provide evidence that penmate genetic effects are important. The fractions of variance due to direct genetic effects and due to penmate genetic effects were 40.0% and 0.1%. Pen effects which may be due to penmate environmental effects are important and should be included in models for genetic evaluation.
Technical Abstract: Records on final weights (kg) of pigs of a University of Nebraska Large White/Landrace composite population were analyzed to estimate effects of penmates on individual growth. Data were weights at approximately 180 days of age of 3,524 pigs in 351 pens (9-11 pigs per pen) farrowed from 1999-2005. Area of each pen was 8.18 square meters. The full model (M1) included fixed effects of contemporary group, sex and line, covariates of age and inbreeding coefficient, and random effects of direct additive genetic (a), genetic penmate (pg) and permanent environmental (pe), pen, litter, and residual effects (e). A derivative-free algorithm was used to obtain REML estimates of variance components with the full model and seven reduced models. When pe was dropped from M1 (M2), estimated pen variance increased slightly. When pen was dropped from M1 (M3), estimated pe variance increased considerably. When pg was dropped from M1 (M4), estimated pen variance increased considerably. When both pen and pe were dropped from M1 (M5), estimates of other variances increased. When pg and pe were dropped from M1 (M6), estimated pen variance increased as with M4. When pg and pen were dropped from M1 (M7), estimated pe variance increased considerably. When pg, pen and pe were dropped from M1 (M8), estimates of direct genetic, litter and residual variances increased, especially the residual variance. Chi-square tests showed that M1 was significantly better than M5 and M8 but was not significantly different from M6. Model 6 was significantly better than M8 (P < 0.005). Under management conditions for this experiment, the conclusion is that the model for genetic evaluation should include litter effects, either pen effects or permanent environmental effects of penmates, and possibly genetic penmate effects, a finding in general agreement with results of studies with different populations and locations. Equations for models with pen effects are much easier to solve than for models with permanent environmental penmate effects.